Tamil Nadu hits bull's eye

How people locked horns with the govt


What is Jallikattu?

Jallikattu is an ancient bull-taming sport that originated in the Southern Indian state of Tamil Nadu over 2,000 years ago . Ancient Tamil texts and cave art indicate that it has been an integral part of the region’s culture. The Tamil term ‘jalli’ means coins and ‘kattu’ means tie. In the sport, a pouch of coins (in ancient times, gold or silver) would be tied to the bull’s horns. The bull would then be let loose and the person who could tame the bull would take home the bounty. In modern times, the pouch has been done away with. Jallikattu is celebrated in rural areas in mid-January to coincide with Pongal, the harvest festival.


What happened to Jallikattu?

In May 2014, the Supreme Court of India, the highest judicial body in the land, imposed a ban on the sport, citing animal cruelty. The case was brought by the Animal Welfare Board and People for Ethical Treatment of Animals (PETA). The ban came into effect in 2015. In 2016, the state was hit by the worst floods in its history and there were no protests. This year, several groups of youngsters converged on Marina beach at Chennai to protest against the Supreme Court's verdict.

The issue is important because the sport helps keep native breeds alive. However, those against the sport argue that the bull is deliberately tortured before releasing it into the arena.

They had all come together through Whatsapp, Facebook and Twiter. As the movement gathered momentum, roads were choked, pressure was piled on the state and eventually after three days, an ordinance was passed allowing the sport to take place while the matter is discussed in parliament. The following analysis discusses people's activity on Twitter during three days of protest - Jan 18, 19 and 20.

protest on the beach
Protest at Marina beach, photo by Jaikumar Vairavan


Marina beach saw over 100,000 people gathering for the protest. Several tweets show that people gave live updates from the scene.


Twitter Analysis

Python was used to extract about 0.7 million tweets over a three-day period from January 18-20. During this time, people congregated on Marina beach and demanded that the Central government raise the ban on Jallikattu. They also tweeted relentlessly from the scene.

For this purpose, this journalist took code from Bhaskar Karambelkar's code to overcome Twtter's API limit and gather a fair-sized sample of tweets quickly. This code implements the Tweepy wrapper, so that needs to be installed before the tweets were extracted.


Who were the top tweeters?

The top tweeters turned out to be bots and a mix of people, including one person in the army (going by his Twitter bio), a software engineer and a student. The software engineer, Satheeshwaran, had tweeted about 1,070 times in the span of three days in the original sample of 700,000 with most of his tweets being retweets.

One of the common traits among the top 10 tweeters was that most of them had far more retweets than tweets. This is an important thing to note -- while people may not be expressing their opinions in their own words, they retweet links and other tweets that they think reinforce their opinions.

icon
Several supporters have changed their profile pictures to this icon


Top tweeters, Tweets/Retweets ratio


Bot-spotting

Of the top 10, three of the accounts were clearly bots -- in fact, two of the bots 'admitted' to being bots. For instance, one of the bots was @RealTimeHack, a 'news' bot which excessively tweets about trending topics. It has a whopping 3.3 million tweets to its credit -- hardly possible for a human.

Another bot which, by the looks of it, had been created exclusively for Jallikattu, was @JallikattuBot. The bot simply kept tweeting out a list of jallikattu-related hashtags at frequent intervals. Interestingly, the bot was shut down and it no longer exists. It had belted out about 4,000 tweets in the span of three days, testing Twitter's API limits to the maximum.


The top retweets

The top tweeters are all celebrities who fall into the broad bracket of sports and entertainment. The most retweeted tweet was by Indian cricketer Virender Sehwag, a prolific tweeter with a sense of humour. While his quips do get him quite a few retweets, one of his tweets in Tamil praising the 'wonderful' people of Tamil Nadu and their peaceful form of protest earned him 26,000 retweets. Another top-10 tweet was by actor Mahesh Babu, who called the people of Tamil Nadu 'bold and fearless'. Journalist Sonia Singh's tweet, noting that there was not a single incident of molestation or misconduct of any kind, was retweeted about 9,000 times. Popular radio jockey Balaji Patturaj's video clip arguing why leather is not banned because cows are killed on an industrial scale was retweeted over 9,000 times.

These instances prove beyond doubt that celebrities have a powerful voice on social media and if they take a stand on an issue, it increases people's support base for the same. This is similar to advertising, where getting the celebrity to endorse a product increases the sales; in this case, the product is replaced by a cause.

The Jallikattu case is an example of celebrity power actually turning into on the ground action. At least two examples fortify this hypothesis: One day before the protests, actor Silambarasan Rajendran uploaded a video on Twitter urging people to block key junctions in protest of the ban and at the same time not block access to medical and emergency services. This tweet was retweeted nearly 10,000 times. As the movement gathered momentum, Silambarasan also created a new hashtag called #JallikattuCare and asked people to use it in case they needed anything like food or medical care during the protest.

Sehwag's tweet praising the 'wonderful' people of Tamil Nadu



Another tweet by an actor that was retweeted thousands of times


icon
Protest at Marina beach, photo by Jaikumar Vairavan



Original content

Original content was less than one might imagine -- Just 7.5% of the sample contained original text, and 5.6% contained content without links. Together, the top 10 tweeters accounted for about 9% of the total tweets in the sample. However, most of the top tweeters too had several retweets on their profile and they had retweeted several actors and other celebrities.

How much content was original?


How many times did people tweet?

An analysis shows that most of the people in the sample tweeted between one and four times.

What words did tweeters use the most?

For this analysis, Python was used to count the occurrence of each word in the tweet sample. This was done after common words like 'and' and 'the' were filtered out. Unsurprisingly, the analysis showed that words like 'Tamil' and 'culture' were most used. There were over a 1,000 mentions of culture, and nearly 2,000 mentions of 'Tamil'.

A deeper look into the tweets showed that several of them appealed to a sense of 'Tamil pride' akin to this piece of research that Donald Trump's supporters were more likely to engage people on several issues on Twitter.

Word cloud: most common words in the sample


Where did tweets come from?

One of the primary difficulties of this piece of analysis is that one cannot analyse every single location. (This is user-entered: so the most creative users do enter 'location: somewhere on the blue dot') However, an initial analysis showed that the bulk of tweets that had entered some location (more than 50%) had came from India. But there were also a lot of activity from the Middle East (Dubai, Bahrain, Abu Dhabi), Singapore and Malaysia. The latter two countries have a strong Tamil population. (10% and 5% respectively). There were also hundreds of tweets from the USA. The tweeters fell under several demographic heads. An initial analysis shows that there were students, corporate workers and white-collar workers.

What % of tweets are geo-enabled?



The bottom line

As University of York's David Beer rightly points out on an LSE blog , "...Perhaps the most telling measure was that of ‘amplification’, which was a measure of the generation of activity in the form of responses and shares. This measure can be used to assess the reaction and activity that individuals can provoke." This is true in the case of the Jallikattu tweets. Most of the tweets were retweets or 'amplifications', and hence can be traced back to tweets from influencers like cricketers or actors.

From this, it follows that it pays for journalists to follow activities of celebrities -- faster one can pick up the thread, further one can go down the amplification road.

Social media is fast becoming the indicator of what people are interested in. "Social media gives you a quick and easy way to gauge audience interest. The amount of likes, comments, and shares a post gets can be telling," writes Jess Archer on the MIT news site. In this instance too, following the tag '#jallikattu' can yield several stories on the subject just by looking through a few hundred tweets.

A WIRED article on why people marched worldwide against Donald Trump shows us that Twitter can be used to find out about the issues and concerns of people. In the article, author Heather Whaling says cites Twitter analytics from Cision and says that abortion and healthcare were major concerns of people around the time of the march. In this particular article, saving an age-old tradition and upholding the rights of Tamilians were the major concerns of people. One limitation of Twitter, however, is that it is limited to people who use it, and hence may not be good for an in-depth analysis.

An interesting phenomenon in the jallikattu case was that though the stakeholders are all bull-owning farmers in the rural districts, (where access to internet is limited) it was urban people who took over Twitter and protest venues in the city -- just another indication that though jallikattu cannot reach large cities, Twitter certainly can.